Composite Materials — Engineering Reference

See also (Tier 3 family index): Composites Taxonomy

1. At a glance

A composite is a material formed from two or more constituent phases that, when combined at the macroscopic scale, deliver a property combination no single phase can match. The dominant engineering class is fibre-reinforced polymer (FRP): high-strength continuous fibres (carbon, glass, aramid) embedded in a polymer matrix (epoxy, polyester, vinyl ester, thermoplastic). Less common but increasingly important are metal-matrix composites (MMC), ceramic-matrix composites (CMC), and particulate composites (filled polymers, cermets, concrete).

The defining feature is anisotropy. Properties vary by direction — a unidirectional carbon/epoxy lamina is roughly an order of magnitude stiffer along the fibre than across it (E_‖ / E_⊥ ≈ 10–15). Engineers exploit this by orienting fibres along load paths and stacking plies into a laminate whose effective ABD-matrix response can be tuned for stiffness, strength, thermal expansion, and even bend-twist coupling.

Why engineers reach for composites. Specific stiffness (E/ρ) and specific strength (σ/ρ) — composites win whenever weight, not absolute strength, drives cost. A T800-grade CFRP unidirectional lamina at V_f = 0.6 stores ≈ 70 MJ/kg of stiffness (about 4× steel and 3× aluminum) and ≈ 3000 kJ/kg of specific strength (4–5× steel). Secondary advantages: directional stiffness tailoring, near-zero longitudinal CTE for satellite optical benches, dielectric transparency (GFRP radomes), corrosion immunity (chemical tanks, marine), and fatigue insensitivity for steel-like loadings.

Where composites dominate the design stack.

• Aerospace primary structure — Boeing 787 (~50 % composite by weight: fuselage barrels, wings, empennage), Airbus A350 (~53 %), F-35, V-22, Eurofighter.
• Wind turbine blades — 80+ m GFRP shells with CFRP spar caps; the single largest volume use of structural composites by tonnage.
• Pressure vessels — Type IV hydrogen tanks (700 bar, CFRP overwrap on polymer liner), SCBA cylinders, CNG tanks.
• Automotive — BMW i3 / i8 CFRP passenger cell; McLaren, Bugatti, Koenigsegg monocoques; carbon-ceramic brakes.
• Sports goods — bicycles, tennis rackets, fishing rods, skis, golf shafts, hockey sticks, racing-shell oars.
• Robotics and UAVs — multirotor airframes, robotic arm links, satellite booms — anywhere mass adds either inertia or launch cost.
• Civil infrastructure — GFRP rebar in chloride environments, FRP bridge decks, filament-wound chemical pipe.

Where composites lose. High raw-material cost (aerospace prepreg $50–200/kg), large recurring tooling investment, brittle damage tolerance (BVID — barely-visible-impact-damage — can knock 30–50 % off compression-after-impact strength), labour-intensive fabrication outside automated aerospace lines, difficulty of inspection and repair, and end-of-life recycling that remains unsolved for thermoset matrices.

---

2. First principles

2.1 What a composite is

A composite has at least two phases:

• Reinforcement — the high-strength, high-stiffness load-bearing phase. In FRP it is the fibre; in MMC the ceramic particle or whisker; in concrete the aggregate.
• Matrix — the continuous binder phase that transfers load between reinforcement units, protects them from abrasion and environment, and sets the geometry. The matrix governs transverse, shear, compression, fatigue, and environmental performance.
• Interface (interphase) — a thin transition zone, often a chemical sizing on the fibre, that controls load transfer and damage initiation. A weak interface gives high toughness via pull-out; a strong interface gives high strength but brittle failure. Tuning the interphase is its own field of materials science.

2.2 Rule of mixtures (Voigt and Reuss bounds)

For a unidirectional lamina with fibre volume fraction V_f and matrix volume fraction V_m = 1 − V_f:

Longitudinal (along the fibre) — Voigt / iso-strain upper bound:

E_‖ = V_f · E_f + V_m · E_m

σ_‖ ≈ V_f · σ_f (when V_f exceeds a small critical fraction; fibre-dominated)

Transverse (across the fibre) — Reuss / iso-stress lower bound:

1 / E_⊥ = V_f / E_f + V_m / E_m

Reuss systematically under-predicts transverse modulus. The Halpin–Tsai semi-empirical model is what engineers actually use:

E_⊥ = E_m · (1 + ξ · η · V_f) / (1 − η · V_f), η = (E_f/E_m − 1) / (E_f/E_m + ξ)

ξ ≈ 2 for transverse modulus of round-fibre composites. Halpin–Tsai also applies to in-plane shear G_12 with ξ ≈ 1.

2.3 Anisotropy ratios

For T800 CFRP UD at V_f = 0.6:

Direction	Modulus	Ratio to E_‖
Longitudinal E_‖	165 GPa	1.0
Transverse E_⊥	8 GPa	0.05
In-plane shear G_12	5 GPa	0.03

This is the engineering reality of unidirectional composites: a structure designed against transverse or shear stress, sized as if it were isotropic, will fail at one tenth of its expected load. Laminate design exists to fix this.

2.4 Quasi-isotropic and balanced layups

Stacking plies at multiple angles produces a laminate whose in-plane stiffness can approximate isotropy:

• [0/±45/90]_s — eight plies, mirror-symmetric. In-plane E_x = E_y; out-of-plane still strongly anisotropic.
• [0/60/120] or [0/±60] — true quasi-isotropic with three angles; less common but mathematically minimal.

A QI CFRP laminate from T800/8552 at V_f = 0.6 gives in-plane E_x ≈ E_y ≈ 50–55 GPa — far less than UD E_‖ but uniform in all in-plane directions. The cost of isotropy is a ~3× drop in achievable stiffness per ply.

2.5 Failure mechanisms are different from metals

Metals fail by ductile yielding or fatigue-driven crack growth in a single phase. Composites have multiple competing damage modes:

• Matrix microcracking — transverse cracks running parallel to fibres. Early, distributed, reduces stiffness ~10 %.
• Fibre-matrix debonding — interphase failure under shear or transverse tension.
• Delamination — interlaminar separation between plies. The signature composite failure mode; often initiates at free edges or impact sites and propagates as crack growth under cyclic load.
• Fibre pull-out — fibres slide out of matrix during failure; primary toughening mechanism in CMC and in well-designed FRP interfaces.
• Fibre kinking and microbuckling — compressive failure where fibres act as columns on an elastic foundation (the matrix). Sets compression strength at ~50–70 % of tensile.
• Fibre tensile rupture — the ultimate strength-of-materials limit, fibre-dominated.

2.6 Classical laminate theory (CLT)

For a thin laminated plate, plane-stress, with N plies each at orientation θ_k:

{N, M}ᵀ = [ABD] · {ε⁰, κ}ᵀ

Where:

• A = in-plane stiffness matrix (units: N/m), A_ij = Σ_k Q̄_ij^(k) · (z_k − z_{k−1})
• B = bending-extension coupling matrix (units: N), zero for symmetric laminates
• D = bending stiffness matrix (units: N·m), D_ij = (1/3) Σ_k Q̄_ij^(k) · (z_k³ − z_{k−1}³)

The ply-level reduced stiffness Q is transformed to the laminate axes via the angle θ_k. CLT is the foundation of every composite-structural-analysis tool (HyperSizer, ESAComp, ABAQUS composite layups, NASTRAN PCOMP/PCOMPG).

2.7 First-ply failure (FPF) vs last-ply failure (LPF)

• FPF — load at which the first ply in any direction fails by any criterion. Conservative, used as design-allowable benchmark.
• LPF — load at which the laminate collapses; usually after several plies have failed and load redistributes. Closer to test data but harder to predict.

Failure criteria for the lamina:

• Maximum stress / maximum strain — fail if any component exceeds its allowable. Simple, non-interactive.
• Tsai–Hill — quadratic, interactive, ignores tension/compression difference.
• Tsai–Wu — quadratic with tension/compression interaction terms. Most popular in industry.
• Hashin / Puck — physics-based, separates fibre and matrix failure modes. Modern preferred for damage-tolerance analysis.

---

3. Practical math / design equations

3.1 Rule of mixtures — worked example 1 (T700S CFRP UD)

Problem. Toray T700S carbon fibre in Hexcel 8552 epoxy matrix, V_f = 0.60. Predict E_‖ and σ_‖,t. Constituent data:

• T700S: E_f = 230 GPa, σ_f = 4900 MPa, ρ_f = 1.80 g/cm³
• 8552 epoxy: E_m = 3.5 GPa, σ_m,y = 80 MPa, ρ_m = 1.30 g/cm³

Longitudinal modulus:

E_‖ = 0.60 · 230 + 0.40 · 3.5 = 138 + 1.4 = 139 GPa (20.2 Msi)

Published value for T700S/8552 UD at V_f = 0.6: E_‖ = 135 GPa. Rule of mixtures over-predicts by ~3 % because real fibres are not perfectly aligned (~1° waviness costs a few percent).

Longitudinal tensile strength:

σ_‖,t ≈ V_f · σ_f = 0.60 · 4900 = 2940 MPa (426 ksi)

Published B-basis allowable per CMH-17: σ_‖,t ≈ 2550 MPa. Rule of mixtures over-predicts by ~13 %; the gap is the fibre strength distribution (Weibull) plus matrix-induced stress concentration at fibre breaks.

Laminate density:

ρ = 0.60 · 1.80 + 0.40 · 1.30 = 1.08 + 0.52 = 1.60 g/cm³

About 4.9× lighter than steel (7.85) at higher stiffness per unit mass.

3.2 Worked example 2 — Quasi-isotropic laminate under in-plane load

Problem. 8-ply T800S/3900-2 QI laminate, [0/+45/−45/90]_s, ply thickness 0.125 mm, total t = 1.0 mm. Applied in-plane axial load N_x = 200 kN/m (200 N/mm). Find ε_x and check FPF.

Step 1. Ply properties (V_f = 0.6):

• E_1 = 165 GPa, E_2 = 8.4 GPa, G_12 = 5.0 GPa, ν_12 = 0.30
• σ_1t = 2750 MPa, σ_1c = 1670 MPa, σ_2t = 64 MPa, σ_2c = 250 MPa, τ_12 = 110 MPa

Step 2. QI in-plane modulus. For a balanced symmetric QI laminate, the in-plane effective modulus is given by:

E_x,QI ≈ (3 · E_1 + 3 · E_2) / 8 + (G_12 / 4) · (some correction)

The standard textbook result for [0/±45/90]_s QI from T800-class CFRP is E_x ≈ 52 GPa.

Step 3. Strain.

ε_x = (N_x / t) / E_x = (200 N/mm / 1 mm) / 52 000 MPa = 200 MPa / 52 000 MPa = 3850 microstrain (0.385 %)

Step 4. FPF check. The 90° plies see the laminate strain transversely — they fail first. Allowable transverse strain for T800S/3900-2 ≈ 0.6 % (8400 microstrain). FPF margin: 0.6 / 0.385 ≈ 1.56 — adequate but not generous. In aerospace practice we add knockdowns for environment (hot-wet ≈ 0.85), impact damage (≈ 0.65), and statistical B-basis (≈ 0.85) — net ≈ 1.56 · 0.47 ≈ 0.73, meaning this layup is under-designed by a factor of 1.4 for an aerospace primary-structure allowable. Either thicken the laminate or rotate plies toward the load.

3.3 Worked example 3 — Netting analysis for a CFRP pressure vessel

Problem. Closed-ended cylindrical pressure vessel, internal pressure p, radius r, wall built from helically wound CFRP at angle ±α to the axis. Find the optimal α that exactly balances hoop and axial loads.

Membrane stresses:

σ_hoop = p·r / t, σ_axial = p·r / (2·t)

Ratio σ_hoop : σ_axial = 2 : 1.

Netting (fibre-only, no matrix) condition: The fibre force component projected onto the hoop direction must equal the hoop stress; the projected axial component must equal axial stress. For a helical winding at ±α:

tan²(α) = σ_hoop / σ_axial = 2

α = arctan(√2) = 54.74° (the so-called “magic angle”)

This is why filament-wound CNG, hydrogen, and SCBA bottles are wound at very close to 55° helical angle, usually with additional pure-hoop overwrap layers at the cylinder mid-section. The 54.74° angle is independent of fibre or matrix system — it follows from the 2:1 stress ratio alone.

3.4 Specific properties — composite advantage chart

Material	ρ (g/cm³)	E (GPa)	σ_u (MPa)	E/ρ (MJ/kg)	σ/ρ (kJ/kg)
Mild steel (A36)	7.85	200	400	25.5	51
4340 Q&T	7.85	200	1500	25.5	191
6061-T6 Al	2.70	69	310	25.5	115
Ti-6Al-4V	4.43	114	950	25.7	215
E-glass/epoxy UD V_f = 0.55	2.0	40	1080	20	540
T700S/8552 CFRP UD V_f = 0.6	1.60	135	2550	84	1594
T800S/3900-2 CFRP UD V_f = 0.6	1.58	165	2750	104	1741
M55J HM CFRP UD V_f = 0.6	1.65	320	1900	194	1152
Kevlar 49/epoxy UD V_f = 0.6	1.38	76	1380	55	1000
CFRP QI [0/±45/90]	1.58	52	750	33	475

UD CFRP wins by 3–7× on specific stiffness against any metal. The QI laminate — what you actually build with — wins by ~1.3× on stiffness and ~5× on strength against mild steel, but the manufacturing cost is 50–500× higher.

---

4. Reference data — common composite systems

System	Form	V_f	E_‖ (GPa)	σ_‖,t (MPa)	σ_‖,c (MPa)	ρ (g/cm³)	Service T (°C)	Use
T300/5208	UD prepreg	0.60	138	1500	1200	1.55	120	First-gen aerospace (1980s)
T700S/2510	UD prepreg	0.55	121	2207	1180	1.55	100	GA aircraft, AGATE allowables
T800S/3900-2	UD prepreg	0.60	165	2750	1670	1.58	121	787, A350, F-35
IM7/8552	UD prepreg	0.58	161	2724	1690	1.58	121	Aerospace baseline
M55J/epoxy	UD prepreg	0.60	320	1900	850	1.65	121	Satellite optical benches
E-glass/epoxy	UD	0.55	39	1080	620	2.0	100	Wind blades, generic FRP
E-glass/vinyl ester	Wet layup	0.40	25	700	480	1.9	80	Marine, tanks
S-glass/epoxy	UD	0.55	53	1750	900	2.0	100	Armour, rocket motor cases
Kevlar 49/epoxy	UD	0.60	76	1380	280	1.38	120	Soft armour, cables, hybrids
UHMWPE (Dyneema)	UD	0.65	90	2600	100	0.97	80	Ballistic armour
AS4/PEEK (APC-2)	UD prepreg	0.60	134	2070	1100	1.57	250	A350 clips, F-22 ducting
CF/PPS	UD	0.55	130	1900	1050	1.55	220	Auto + aero brackets
SiC/SiC CMC	2D weave	n/a	200	380	n/a	2.7	1300	LEAP engine shrouds
SiC-particle/6061 Al MMC	particulate	0.20	100	480	480	2.85	250	Brake rotors, missile fins

Properties per ASTM D3039 (longitudinal tensile), ASTM D3410 or D6641 (longitudinal compression). Tabulated values are typical test means; design allowables are typically B-basis (CMH-17) ≈ 65–75 % of mean for fibre-direction tension, lower for matrix-dominated and compression.

---

5m. Composition & microstructure

5m.1 Carbon fibres

The dominant aerospace and high-performance fibre. Made by thermal pyrolysis of a polymeric precursor — either PAN (polyacrylonitrile) for the structural grades or pitch (mesophase) for ultra-high-modulus and thermal-management fibres.

Toray standard-modulus (SM) and intermediate-modulus (IM) grades — PAN-based, the workhorses:

Grade	E_f (GPa)	σ_f,u (MPa)	ε_u (%)	ρ_f (g/cm³)	Comment
T300	230	3530	1.5	1.76	First-gen, low cost, declining
T700S	230	4900	2.1	1.80	Industrial workhorse
T800S	294	5880	2.0	1.80	Aerospace baseline (787, A350)
T1000G	294	6370	2.2	1.80	High-strength specialty
T1100G	324	7000	2.0	1.79	Latest IM-class

High-modulus (HM) pitch-based fibres — Mitsubishi K-series, Toray M-series:

Grade	E_f (GPa)	σ_f,u (MPa)	ε_u (%)	Use
M40J	377	4400	1.2	Satellite booms
M55J	540	4020	0.8	Optical benches, near-zero CTE
M60J	588	3820	0.7	Highest-modulus PAN
K1100 (pitch)	931	3100	0.3	Thermal radiators (k = 1100 W/m·K)

Suppliers and market. Toray dominates (~60 % world capacity including its Toray-Hexcel and ZOLTEK lines); Hexcel (US), SGL Carbon (DE), Mitsubishi Chemical, Solvay (former Cytec), Teijin (Tenax). Industrial-grade large-tow (50K filaments) fibres at ~$15–25/kgaredrivingwindbladeandautomotiveadoption;aerospacesmall-tow(3K,6K,12K)prepregsitsat$50–200/kg.

5m.2 Glass fibres

Cheap, ubiquitous, and the volume leader in FRP by tonnage.

Type	Composition family	E_f (GPa)	σ_f,u (MPa)	ρ_f	Use
E-glass	Calcium aluminoborosilicate	72	3450	2.55	Standard FRP, electrical
S-glass / S-2	Mg-aluminosilicate	86	4750	2.46	Ballistic, motor cases
R-glass	Ca-Al-silicate	86	4400	2.55	EU equivalent of S-glass
AR-glass	Zirconia-modified	72	3000	2.70	Alkali-resistant; concrete reinforcement
C-glass	Soda-lime-borosilicate	69	3300	2.50	Chemical-resistant veils

Suppliers: Owens Corning, Jushi (CN, world’s largest by volume), 3B Fibreglass, Nippon Electric Glass. Bulk E-glass roving runs $2–5/kg — the reason E-glass owns wind, marine, and tanks.

5m.3 Aramid fibres

Para-aramid (poly(p-phenylene terephthalamide)): the highly oriented, hydrogen-bonded liquid-crystal polymer.

• Kevlar 49 (DuPont) — E_f = 124 GPa, σ_u = 3620 MPa, ε_u = 2.4 %, ρ = 1.44 g/cm³. The standard structural aramid.
• Kevlar 29 — lower modulus (70 GPa), higher elongation (3.5 %) — for soft armour, ropes, cables.
• Kevlar 149 / KM2 / Plus — higher-modulus variants (E_f = 143–186 GPa) for aerospace and armour upgrades.
• Twaron (Teijin) — para-aramid equivalent, similar property class.
• Technora (Teijin) — copolymer aramid, better fatigue and chemical resistance.

Aramids are unmatched for impact, cut, and ballistic performance per unit mass; they are useless in compression (σ_‖,c ≈ 0.2 × σ_‖,t — the fibre kinks at low load). UV degrades aramid; bare yarn loses 50 % strength in 1 year of sun exposure. Machining aramid composites is notorious — the tough fibres fuzz rather than cut; use serrated or scissor-action tooling.

5m.4 UHMWPE fibres

Ultra-high-molecular-weight polyethylene gel-spun fibre — Dyneema (DSM, now Avient) and Spectra (Honeywell).

• E_f = 100–130 GPa, σ_u = 2400–3500 MPa, ρ = 0.97 g/cm³ (only material that floats).
• Lowest density of any structural fibre — best ballistic-performance per kg.
• Service ceiling ~80 °C (polyethylene matrix melts at 130 °C); creeps under sustained load; cannot be used above 60 °C structural.

5m.5 Other fibres

• Basalt — drawn from molten volcanic rock; properties similar to S-glass; higher T service (~600 °C continuous); marketed as sustainable. Suppliers: Kamenny Vek (RU), Mafic, Sudaglass.
• Boron fibre — CVD boron on tungsten wire core; E_f = 400 GPa, σ_u = 3600 MPa, ρ = 2.5 g/cm³. Used in F-15, B-1, and patches for aging-aircraft repair; very expensive (~$1000/kg).
• SiC fibre — Nicalon, Tyranno (Japanese). Ceramic, for CMC matrices; E_f = 200–400 GPa.
• Natural fibres — flax, hemp, jute, bamboo. Specialty automotive interior panels, low-grade composites; sustainability focus.

5m.6 Polymer matrices

Family	Trade names	T_g (°C)	T_service	Use share	Comment
Epoxy	Hexcel 8552, Cycom 977-3, Toray 3900-2, Newport NCT-301	120–200	RT–177	~85 % of aerospace primary	Workhorse thermoset
Unsaturated polyester	many generic	60–110	RT–80	High volume marine/auto	Cheap, wet layup, lower toughness
Vinyl ester	Derakane, Atlac	100–130	RT–110	Tanks, marine	Better chemical resistance than polyester
Phenolic	many	n/a (cures to char)	RT–150	FAA cabin interior (FAR 25.853 fire)	Low smoke/toxicity; brittle
Cyanate ester	Bryte EX-1515, Cycom 5575-2	250–290	RT–230	Satellite, radomes	Very low CTE, low outgassing
BMI (bismaleimide)	Cycom 5250-4, Hexcel HexPly F655	250–290	RT–230	F-35 wing skins, engine cowls	High-T thermoset
PI (polyimide)	PMR-15, RM-1100	320–370	RT–315	Engine compressor stators	Hardest to process; toxic precursors
PEEK	Victrex APC-2, Solvay KetaSpire	143	RT–250	A350 clips, F-22 ducting, medical	Thermoplastic; weldable; recyclable
PPS	Toray Cetex	90	RT–220	Automotive structural, A380 fittings	Thermoplastic; lower cost than PEEK
PEI (Ultem)	SABIC	217	RT–170	Aerospace clips and ducting	Amorphous thermoplastic; transparent
PEKK	Arkema Kepstan	156	RT–250	A350, V-280 brackets	PEEK rival, lower melt T

Why epoxy dominates aerospace. Cure window matches autoclave technology (177 °C / 6 bar), interlaminar fracture toughness G_IC ≈ 200–600 J/m² (with toughened systems), bonding chemistry well understood, half a century of allowables database. Toughened epoxies (e.g. Hexcel 8552 with thermoplastic CTBN/PES toughener) reach G_IC > 500 J/m², approaching thermoplastic ductility while retaining thermoset processability.

5m.7 Metal-matrix composites (MMC)

• Al + SiC particles (10–30 vol %) — Duralcan, AMC composites. σ_u + E ≈ 30–50 % above base alloy. Used in Toyota brake rotors, missile fins, brake calipers (BMW E36 M3), bicycle frames historically. Cost roughly 5–10× base aluminum.
• Al + Al₂O₃ fibre (Saffil) — DiCV (discontinuous fibre); engine piston ring grooves (Toyota, Honda diesel pistons).
• Ti + SiC fibre (TMC) — SCS-6 / Ti-6Al-4V matrix. Compressor blings (bladed rings) in advanced engine programs.
• Cu + W or Cu + diamond — electronic-package heat spreaders.

MMC properties depend heavily on reinforcement geometry (continuous fibre, whisker, particulate) and processing (powder-metallurgy, liquid-state infiltration, in-situ).

5m.8 Ceramic-matrix composites (CMC)

• SiC/SiC — SiC fibre tows in a SiC matrix densified by chemical-vapour infiltration (CVI), polymer infiltration and pyrolysis (PIP), or melt infiltration (MI). Continuous service to 1200 °C in oxidising atmosphere, 1300 °C in inert. GE Aviation LEAP engine HPT shrouds are SiC/SiC — first commercial flight on LEAP-1A (A320neo) in 2016. ~30 % weight reduction vs Ni-base superalloy plus eliminates cooling air bleed.
• C/C (carbon-carbon) — pyrolytic-carbon-matrix in carbon fibre. Brake discs (F1, military aircraft, Boeing 787 brakes), rocket nozzle throats, Shuttle wing leading edges. Oxidises catastrophically above 500 °C in air — needs SiC coating for re-usable systems.
• C/SiC and Cf/SiC — hybrid; better oxidation than C/C, higher toughness than SiC/SiC. Used in carbon-ceramic brake discs (Porsche PCCB, Ferrari).
• Oxide-oxide (Al₂O₃ fibre + Al₂O₃ matrix) — 3M Nextel-based; lower-T (1100 °C) but oxidation-immune by chemistry; emerging in commercial engine exhaust components.

5m.9 Particulate composites

• Concrete — coarse + fine aggregate in hydrated Portland cement paste matrix. Highest-tonnage composite on Earth. (See [[Engineering/reinforced-concrete]].)
• Filled engineering polymers — talc/CaCO₃/glass-bead in PA66, PP, PBT to stiffen, control CTE, reduce cost. Already covered in [[Engineering/materials-polymers]].
• Tungsten carbide-cobalt (WC-Co cermet) — 6–30 % Co binder around WC grains. Cutting tool inserts (Sandvik, Kennametal), drill bits, wear parts. Hardness 1400–1800 HV, fracture toughness 8–20 MPa√m — far tougher than monolithic WC.
• Cermet thermal-barrier coatings — yttria-stabilised zirconia in a NiCrAlY bond coat on turbine blades.

5m.10 Layup notation

Stacking sequence is written ply-by-ply from one face inward, separated by slashes:

• [0/90/+45/−45] — 4-ply non-symmetric.
• [0/+45/−45/90]_s — 8-ply symmetric (subscript s = mirror about midplane).
• [0/±45/90]_s — same as above with compact notation for ±45 pair.
• [0_2/90_2]_s — two adjacent 0° plies, two 90° plies, mirrored: 8-ply total.
• [(0/90)_3]_s — three repeats of (0/90) then mirrored: 12-ply total.
• Subscript T denotes total (non-mirrored) count.

Symmetric and balanced layups (every +θ has a matching −θ at the same distance from midplane) zero out the B matrix → no bend-extension coupling → flat parts after cure.

---

6m. Mechanical properties

6m.1 UD lamina baseline data (V_f = 0.6 prepreg, room temperature, dry)

System	E_‖ (GPa)	E_⊥ (GPa)	G_12 (GPa)	ν_12	σ_‖,t (MPa)	σ_‖,c (MPa)	σ_⊥,t (MPa)	σ_⊥,c (MPa)	τ_12 (MPa)
T300/5208 (legacy)	135	10.0	5.0	0.30	1500	1200	50	200	70
T700S/2510 (AGATE)	121	8.6	4.4	0.30	2207	1180	53	200	90
T800S/3900-2	165	8.4	5.0	0.32	2750	1670	64	250	110
IM7/8552	161	11.4	5.2	0.32	2724	1690	64	286	120
M55J/epoxy	320	6.5	4.0	0.34	1900	850	30	130	70
E-glass/epoxy	39	8.5	4.0	0.28	1080	620	39	130	70
Kevlar 49/epoxy	76	5.5	2.3	0.34	1380	280	30	130	60

All per ASTM D3039 (tension) and ASTM D3410 or D6641 (compression). ASTM D3518 for in-plane shear via [±45]_ns coupon. ISO 14130 or ASTM D2344 for short-beam interlaminar shear strength (ILSS ≈ 70–100 MPa typical aerospace CFRP).

6m.2 Compression and the fibre-microbuckling limit

Compressive strength σ_‖,c is always less than tensile σ_‖,t for fibre composites — typically 50–70 %. The mechanism is fibre microbuckling: each fibre acts as a column on an elastic foundation (the matrix). The Rosen-Argon buckling model predicts:

σ_‖,c ≈ G_12 / (1 − V_f)

For T800/3900-2: G_12 ≈ 5 GPa, V_f = 0.6 → σ_‖,c ≈ 12 500 MPa. Real strength is 1670 MPa — about an order of magnitude lower because fibre misalignment, voids, and matrix nonlinearity all degrade the elastic-foundation assumption. Compression strength is matrix- and quality-controlled, not fibre-controlled. This is the central engineering reality: improving fibre strength alone does little for compression-driven design.

6m.3 Laminate stiffness — QI baseline

Quasi-isotropic [0/±45/90]_s laminate stiffness for aerospace CFRP at V_f = 0.6:

System	E_x,QI (GPa)	σ_x,QI,t (MPa, OHT)	σ_x,QI,c (MPa, OHC)
T300/5208	47	380	280
T800S/3900-2	56	470	380
IM7/8552	55	450	370
M55J/epoxy	115	250	200
E-glass/epoxy	16	280	240

OHT / OHC = open-hole tension / compression per ASTM D5766 / D6484. These are the values that drive joint and notched-region sizing — the open hole knocks tensile strength to ~30 % of unnotched QI, far more severe than the K_t ≈ 3 of a metal. Composites are notch-sensitive — bolt-hole layups need local doublers or reinforcing plies.

6m.4 Fatigue

Fibre-direction tension-tension fatigue of CFRP is extraordinarily good — the S-N curve drops only ~10 % per decade of cycles, and 10⁷-cycle endurance is typically ~80 % of static σ_‖,t. Steel-grade fatigue endurance is matched at one-fifth the weight, which is the engineering case for CFRP rotor shafts, drive shafts, and pressure-vessel liners.

Matrix-dominated fatigue (transverse and shear) is poor — knockdown to ~30 % static at 10⁷ cycles. Compression-compression and reverse-axial fatigue (R = −1) likewise dominated by matrix damage; CFRP loses 50–60 % static σ_‖,c at 10⁶ cycles.

6m.5 Design allowables

Test data are scattered; structural design uses statistical-basis allowables:

• A-basis — 99 % of population exceeds the value, with 95 % confidence. Single-load-path primary structure.
• B-basis — 90 % exceed, 95 % confidence. Standard for redundant aerospace structure.
• S-basis — specification minimum (vendor-quoted, not statistical).

CMH-17 provides the canonical B-basis database for the aerospace-qualified systems above. AGATE allowables (Advanced General Aviation Transport Experiments, late 1990s) opened B-basis qualification to small-aircraft developers for T700/2510 and a handful of other low-cost systems.

6m.6 Hot-wet knockdowns

Environmental exposure degrades matrix-dominated properties:

• Moisture absorption 1–2 % by mass at 95 % RH equilibrium (typically 1000–5000 hours).
• T_g_wet ≈ T_g_dry − 30 °C (the well-known plasticisation effect).
• σ_‖,c drops 20–30 % hot-wet (typically 82 °C / saturated).
• σ_⊥,t drops 20–40 % hot-wet.
• E_‖ essentially unaffected (fibre-dominated).

Aerospace design uses ETW (elevated-temperature wet) allowables, typically at the maximum operating-temperature minus the airframe T_g margin (~28 °C / 50 °F per FAA AC 20-107B).

---

7m. Thermal / electrical / chemical properties

7m.1 Density

System	ρ (g/cm³)	vs steel (7.85)	vs Al (2.70)
CFRP UD (T800/8552)	1.58	0.20	0.59
GFRP UD (E-glass/epoxy)	2.00	0.25	0.74
AFRP UD (Kevlar 49/epoxy)	1.38	0.18	0.51
UHMWPE/epoxy	1.05	0.13	0.39
CFRP QI laminate	1.58	0.20	0.59

7m.2 Coefficient of thermal expansion (CTE)

CFRP is among the few materials with negative longitudinal CTE (the carbon-fibre crystal lattice contracts when heated):

System	α_‖ (10⁻⁶ /°C)	α_⊥ (10⁻⁶ /°C)	α_QI (10⁻⁶ /°C)
T800/3900-2 CFRP	−0.5	30	2.5
M55J HM CFRP	−1.1	28	0.3
E-glass/epoxy	7.0	30	14
Kevlar 49/epoxy	−4.0	60	5
Steel	12	12	12
6061 Al	24	24	24

Near-zero CTE laminates are a key satellite-structure capability — optical benches and telescope tube structures are routinely designed to α ≈ 0.1 × 10⁻⁶ /°C by laminate balancing. The James Webb Space Telescope optical bench is a CFRP sandwich.

CTE mismatch problems — bolting CFRP to aluminum across a 100 °C temperature range generates ~2.2 mm/m of differential expansion. Slot the holes, use spring elements, or accept the resulting joint preload variation.

7m.3 Thermal conductivity

System	k_‖ (W/m·K)	k_⊥ (W/m·K)
Standard PAN CFRP (T800)	7	0.8
K1100 pitch CFRP	600	1
GFRP	0.3	0.3
AFRP	0.04	0.04

PAN-based CFRP is moderately conductive along fibre (carbon’s basal-plane conductivity dominates). Pitch-based CFRP (K1100) approaches copper conductivity along fibre, while remaining a transverse insulator — the basis for spacecraft thermal radiators and heat-sink substrates. GFRP is an insulator both ways; AFRP is the best polymeric thermal insulator of structural materials.

7m.4 Electrical properties

• CFRP — electrically conductive along fibre (~10⁻⁵ Ω·m), nearly insulating across (10⁰–10² Ω·m). Conducts lightning along the lay-up, will galvanically corrode any less-noble metal it contacts.
• GFRP — full dielectric. Volume resistivity 10¹³–10¹⁵ Ω·cm. Radomes, motor end-bells, switchgear, GFRP rebar electrical isolation.
• AFRP — dielectric similar to GFRP.

Galvanic corrosion at CFRP-metal joints. CFRP is electrochemically noble (sits near platinum on the galvanic series in seawater). Bolt CFRP to aluminum and the aluminum dissolves at the interface within months in marine atmosphere. Mitigation: glass-ply isolation layer (a single 0.1 mm ply of GFRP between CFRP and Al), sealant (PRC PR-1422, MIL-PRF-81733 Type IV) on faying surface, sealed fastener installation. Steel is less aggressive but still affected. Titanium is electrochemically compatible with CFRP — preferred fastener material in 787 / A350 wing-to-fuselage joints.

7m.5 Chemical resistance

• Epoxy matrix — resists most chemicals; attacked by strong acids, ketones (MEK, acetone), and chlorinated solvents. Hot oils swell over years.
• Vinyl ester — superior to epoxy in dilute acids and chlorine-bearing fluids; preferred for FGD scrubbers and chlor-alkali tanks.
• Polyester — adequate for water, fuel, dilute salt; degrades in strong acids, alkalis, and many solvents.
• Phenolic — excellent acid resistance and fire performance; degrades in strong alkalis.
• PEEK matrix — outstanding chemistry resistance, including hot oil, fuels, and hydraulic fluids. Attacked only by concentrated sulfuric acid.
• Fibres — carbon is chemically inert below 500 °C in air; glass is attacked by HF and concentrated NaOH; aramid is hydrolysed in hot acid.

7m.6 Moisture and hygrothermal effects

• Epoxy absorbs 1–2 % water by mass over years at 50–95 % RH; equilibration time τ ≈ L²/D, D ~10⁻⁷ mm²/s at 50 °C, so a 5 mm laminate equilibrates in ~2000 hours.
• T_g drops 25–35 °C from dry to saturated.
• Plastic-shear strength σ_⊥,t and τ_12 drop 20–40 %.
• Reversible: drying restores ~95 % of properties.

Aerospace certification testing must include ETW (Elevated-Temperature Wet) coupons aged to saturation per ASTM D5229.

7m.7 Radiation and UV

• Carbon fibre is radiation-immune; satellite primary structure routinely accumulates 100 kGy without measurable property change.
• Epoxy matrix absorbs UV (yellows, surface degrades); requires paint or coating outdoors.
• Aramid UV-degrades within months bare; armour panels coated or jacketed.
• Gamma radiation embrittles polymer matrices at doses >10 kGy; affects nuclear-cell composites significantly.

7m.8 Lightning strike

A carbon-aircraft must conduct lightning across its skin without burning through. Boeing 787 and Airbus A350 employ expanded copper foil (ECF) or bronze mesh integrated into the outermost ply, plus conductive sealants and metal-foil fastener heads at lightning attach points. Per FAA AC 20-53B and SAE ARP5412, primary structure must survive a 200 kA Zone 1A strike without structural perforation. Test labs (Lightning Technologies, ENSTA Bretagne) deliver up to 2 MV / 200 kA full-vehicle attach tests.

---

8m. Processing & joining

8m.1 Manufacturing processes

Process	V_f range	Void content	Cost class	Application
Hand wet layup	0.30–0.45	2–5 %	Lowest	Boats, repairs, prototypes
Vacuum-bag wet layup	0.40–0.55	1–3 %	Low	Boats, wind blades, custom
VARTM / SCRIMP	0.50–0.60	< 1 %	Low-medium	Wind blades, military hulls
RTM (resin transfer molding)	0.55–0.65	< 1 %	Medium	Automotive, A350 spars, brackets
Prepreg + autoclave	0.55–0.62	< 0.5 %	High	Aerospace primary structure
Out-of-autoclave (OOA) prepreg	0.55–0.60	< 1 %	Medium-high	Modern alt to autoclave
Compression molding (SMC/BMC)	0.20–0.50	1–2 %	Medium (high-vol)	Automotive Class A, electrical
Filament winding	0.60–0.68	< 1 %	Medium	Pressure vessels, pipe, drive shafts
Pultrusion	0.60–0.70	< 1 %	Medium (high-vol)	Constant-section profiles
AFP / ATL	0.55–0.60	< 0.5 %	Very high (capital)	787 fuselage, A350 wings
Thermoplastic stamping	0.55–0.60	< 1 %	Medium-high	A350 ribs, autoclips
Continuous-fibre 3D print	0.30–0.40	2–10 %	Medium	Markforged, Arevo, Anisoprint

Autoclave = pressure vessel up to 6–7 bar internal, heated to 121–180 °C, holds part on tool under vacuum bag. Standard aerospace prepreg cure cycle: 1.7 °C/min ramp, 2 h hold at 177 °C, 6 bar autoclave pressure, controlled cooldown. Boeing and Airbus operate autoclaves up to 9 m diameter and 30 m length — capital cost $30–80M each.

AFP = Automated Fibre Placement. Robotic head lays 1/4” or 1/2” prepreg slit tape along programmed paths, building up complex contours ply by ply. 787 fuselage section barrels are one-piece AFP layups on internally heated mandrels. Capital cost $20–40M per machine plus tooling.

8m.2 Joining

Method	Pros	Cons	Use
Mechanical fasteners (bolts, rivets)	Removable, inspectable, well-understood	Stress concentration, delamination, weight	Aerospace primary structure
Adhesive bonding	Distributes load, no holes, low weight	Surface prep critical, NDI difficult, no inspection at runtime	Skin-to-substructure, repair
Co-curing	Single cure cycle, monolithic	Tooling complexity	Stiffened panels, integral structures
Co-bonding	Bond cured to uncured part	Surface prep, peel-ply transfer	T-joints, lap joints
Thermoplastic welding (induction, resistance, ultrasonic)	Fast, recyclable, in-process inspection	Limited to TP matrices	A350 thermoplastic clips, V-280

Mechanical fastening rules:

• Edge distance ≥ 3·d (vs 1.5·d for metals).
• Hole diameter / thickness > 1 for clean drilling.
• Use titanium fasteners in CFRP (electrochemically compatible); steel and aluminum cause galvanic problems.
• Bearing strength scales with laminate quasi-isotropy and matrix toughness; B-basis bearing typically 600–800 MPa for aerospace CFRP per ASTM D5961.
• Hi-Lock fasteners (Hi-Shear / LISI) and threaded rivets (Cherry Aerospace) are aerospace standard; preload-controlled.

Adhesive bonding — epoxy paste (Hysol EA 9394, Henkel EA 9396), epoxy film (Cytec FM 300, 3M AF-163), or PU. Surface prep is everything:

• Peel ply removal — co-cured fabric ply pulled off just before bonding leaves textured, contamination-free surface.
• Plasma treatment — for thermoplastics; raises surface energy.
• Grit blast + solvent wipe — for cured composites and metals.
• Bond-line thickness 0.13–0.25 mm typical; adhesive starvation (too thin) and porosity (too thick) both reduce strength.

8m.3 Machining

Composites tear, delaminate, and abrade tooling. Rules differ markedly from metals:

• Drilling — Use PCD (polycrystalline diamond) or diamond-coated carbide tooling. Brad-point or dagger geometry. Peck-drill (~0.5–1 mm peck) and use a backing plate to prevent exit-side breakout. Aerospace spec: hole quality per NAS 9100 or company internal.
• Routing / trimming — diamond-coated burrs; high RPM (10000–20000) and moderate feed; vacuum dust extraction (carbon dust is conductive — shorts electronics — and a known respiratory hazard at micron-scale).
• Waterjet — abrasive waterjet for trim, no thermal effects, no tool wear. Standard for production trim of large parts.
• Laser cutting — for thin composites only; matrix burns and chars.
• Ultrasonic-assisted machining — emerging for difficult-to-cut aramid; serrated rotary cutters required.

8m.4 NDI / inspection

Composite defects (porosity, delamination, fibre wrinkle, foreign object) are not visible. Inspection is the bottleneck:

• Ultrasonic C-scan — pulse-echo or through-transmission; primary inspection at part level for aerospace. Detects porosity > 1 %, delaminations > 6 mm diameter typical.
• Phased-array ultrasonic — for in-service inspection of complex curvatures.
• X-ray and CT — high-resolution detection of porosity, fibre orientation, ply wrinkles. CT scan of a 1 m³ aerospace part takes days.
• Thermography — IR camera, flash-heating; detects near-surface delaminations rapidly.
• Tap test — coin or specialised hammer; trained ear hears dull “thud” for delamination vs sharp “click” for sound. Cheap field check.
• Shearography — laser interferometry; detects sub-surface defects under load.
• Embedded fibre-optic sensors (FBG) — structural health monitoring; permanent strain history.

8m.5 Repair

Composite repair is heavily regulated; aircraft repair follows the SRM (Structural Repair Manual) and FAA AC 43.13-1B:

• Cosmetic / non-structural — local resin injection, fairing in epoxy filler.
• Bonded scarf repair — taper-grind a long, shallow (1:30 to 1:60 ratio) scarf into damage, lay in matching prepreg plies, cure with heat blanket and vacuum. Standard primary-structure repair.
• Bolted patch repair — titanium or composite doubler bolted across damage; field-expedient or non-primary structure.
• Hot bonder — portable closed-loop cure controller with vacuum bag and thermocouples; the airline shop-floor repair tool.

---

9m. Applications & selection trade-offs

9m.1 Quick-pick guide

Need	First pick	Reason
Weight-critical aerospace primary structure	T800S/3900-2 or IM7/8552 CFRP	Best balance of allowables, qualification, supply chain
Wind turbine blade shell	E-glass/epoxy infusion	Cost-driven; CFRP only in spar caps
Hydrogen pressure vessel (Type IV, 700 bar)	T700S/T800 CFRP filament wound on HDPE liner	Specific strength, fatigue, no leak path
Bicycle frame	T700/T800 CFRP layup	Weight + stiffness; tooling amortised at small volumes
Marine hull	E-glass/vinyl ester	Cost, corrosion, repairability
Chemical tank	E-glass/vinyl ester wound	Corrosion immunity, in-service life
Ballistic armour panel	Kevlar/UHMWPE laminate, S-glass/phenolic	Impact, low density
Radome / antenna housing	Quartz or E-glass / cyanate ester	Low dielectric loss, low T variation
Satellite optical bench	M55J/cyanate ester	Near-zero CTE, dimensional stability
Engine HPT shroud (1200 °C)	SiC/SiC CMC	Temperature beyond Ni superalloy capability
Brake disc, high-perf auto	C/SiC carbon-ceramic	Mass reduction, fade resistance
GFRP rebar in concrete (chloride exposure)	E-CR or basalt FRP rebar	Eliminates corrosion-driven service-life limit
Robotic arm link (collaborative or fast)	T700 CFRP tube + Al end-fittings	Stiffness + inertia reduction
Drone airframe	Plain-weave CF + epoxy or pre-cured tubes	Off-the-shelf supply, repair

9m.2 Trade-offs

• CFRP vs aluminum (aerospace). CFRP is ~30–50 % lighter at equivalent stiffness, but raw material is 10–20× more expensive and manufacturing is 3–5× more labour-hours. The crossover happens where mass directly drives operating cost or mission cost — long-range commercial aircraft, satellites, racing cars, performance UAVs. For ground vehicles where fuel saving alone must justify mass reduction, the breakeven is harder; CFRP dominates supercars where weight is part of the brand, not the BMW 5-Series where it isn’t.

• GFRP vs steel (industrial). GFRP wins on corrosion (no painting in chloride / acid service), electrical isolation, and lower-volume runs (no tooling amortisation), at the cost of stiffness and joint complexity. Tanks, pipes, gratings, ladders, electrical poles, and architectural panels are GFRP markets.

• Hand layup vs prepreg/autoclave (cost vs quality). Hand layup V_f ≈ 0.40 with 2–5 % voids is the hobby and repair end; autoclave V_f 0.60 with < 0.5 % voids is the aerospace end. Strength differs by ~2× and stiffness by ~30 % for the same fibre/matrix system — porosity and resin-rich pockets kill compression and matrix-dominated allowables.

• Thermoset vs thermoplastic matrix. Thermoset (epoxy/BMI/cyanate ester) dominates by volume and qualification heritage. Thermoplastic (PEEK/PPS/PEKK) gains on recyclability, weldability, longer shelf life, faster cycle time (no chemical cure) — Airbus A350 thermoplastic clips and brackets are the leading edge; Boeing 777X has thermoplastic empennage components. Adoption is accelerating in 2020s. Cost remains 2–4× thermoset prepreg.

• AFP/ATL vs hand layup (aerospace). Automation is justified for >5000 part-hours per year per pattern. Below that, skilled hand layup is more economical. AFP wins on consistency, traceability, and waste reduction (zero ply trimming).

9m.3 Notable structural examples

• Boeing 787 Dreamliner — composite fuselage barrels, wings, vertical and horizontal stabilisers; ~50 % composite by weight, 80 % by volume. T800S/3900-2 base material. Entered service 2011.
• Airbus A350 XWB — composite fuselage (panel construction, not barrels), wings; ~53 % composite. IM7/8552 primary baseline.
• Lockheed Martin F-35 — significant CFRP and BMI primary structure; thermoplastic in F-35B doors.
• Boeing 777X — composite folding wingtip and stiffened panel wings; thermoplastic empennage skin.
• BMW i3 / i8 — CFRP passenger cell (Life Module); Tier 1 high-volume CFRP manufacturing (Moses Lake, WA fibre + Wackersdorf molding).
• Toyota Mirai II — Type IV CFRP 70 MPa hydrogen tank, filament-wound T700-class.
• Vestas / GE / Siemens-Gamesa wind blades — 60–120 m blades, GFRP shell with CFRP pultruded spar caps in modern designs.
• JWST optical bench — graphite-cyanate-ester sandwich, near-zero CTE.

---

10m. Failure modes

1. Fibre-direction tensile rupture. Strength-limited, fibre-controlled, brittle. Failure strain ~1.5–2 %. Acoustic emission warning prior to failure but no visible yield. The intended failure mode of fibre-aligned designs.

2. Matrix cracking (transverse). Cracks parallel to fibres in off-axis plies under transverse tension or shear. Reduces laminate stiffness ~10 % at first-ply failure, opens diffusion paths for moisture, initiates delamination. Not catastrophic alone but the trigger for downstream damage.

3. Delamination. Interlaminar separation between plies. Driving forces: out-of-plane (impact, edge), free-edge interlaminar shear, ply-drop discontinuities, post-curing residual stresses. Characterised by G_IC (mode I) and G_IIC (mode II) interlaminar fracture toughness per ASTM D5528 and D7905; mixed-mode per ASTM D6671. Aerospace toughened epoxies achieve G_IC ≈ 400–600 J/m²; PEEK matrix systems achieve G_IC > 1500 J/m². Delamination is the signature composite failure mode and drives damage-tolerance design.

4. Compression failure — fibre microbuckling / kink banding. Under compression, fibres act as columns on the matrix as elastic foundation. Initial fibre misalignment (~1–3°) amplifies under load and the fibres locally buckle into shear-dominated kink bands at ~15–25° rotation. Compression strength is therefore matrix-toughness-controlled and quality-controlled. Wet hot CFRP loses 25 % compression strength.

5. Impact damage and compression after impact (CAI). A 6 J impact on a 4 mm CFRP plate leaves no visible surface mark (BVID — barely-visible-impact-damage) but produces a ~25 mm diameter internal delamination. Compression strength after that impact (CAI per ASTM D7137) drops by 30–50 %. Sets the design knockdown for aerospace primary structure. CMH-17 specifies the test methodology.

6. Notch sensitivity at holes and cutouts. Stress concentration K_t for a CFRP QI laminate with a circular open hole is 3–7, much higher than the K_t ≈ 3 of metals because composites lack plastic blunting at the hole edge. Open-hole tension (OHT) per ASTM D5766 and open-hole compression (OHC) per ASTM D6484 are required allowables. Filled holes (bolt installed) recover some strength via load transfer.

7. Edge effects. Interlaminar normal and shear stresses peak at free edges due to mismatch in Poisson ratio between adjacent plies. Free-edge delamination is a common test-coupon failure that doesn’t always represent real structure; engineers wrap edges or taper-stop plies inboard to suppress it.

8. Hygrothermal degradation. Moisture absorption + temperature reduces matrix-dominated properties as documented above. Reversible. Aerospace allowables are post-conditioning ETW.

9. Galvanic corrosion at CFRP-to-aluminum interfaces. CFRP is electrochemically noble; aluminum dissolves. Mitigation: GFRP isolation ply, sealant, titanium fasteners, faying-surface barrier.

10. Lightning strike damage. Direct attach burns matrix locally; current spread through carbon-fibre paths heats them resistively; can blow plies apart through Joule-heated steam from absorbed moisture. Mitigation: conductive mesh in outermost ply, metallised fastener heads, sealed joints.

11. Long-term creep and stress rupture. Matrix creeps under sustained load; fibres can stress-rupture under long-term tension at 60–80 % of static σ_u. Pressure vessels and rotor blades use sustained-stress allowables typically capped at 35 % of σ_‖,u.

12. Manufacturing defects. Porosity (entrained air, volatiles), fibre waviness (out-of-plane wrinkles), resin-rich pockets, foreign-object inclusions (release film, peel-ply scraps), and delamination from cure-cycle missteps. All knock compression and shear strength faster than tension. CMH-17 Vol. 3 catalogues allowable-vs-defect-size data.

13. Damage tolerance philosophy. Unlike metals (designed “leak before break” or with crack-growth limits between inspections), composites are designed for “no growth” between inspections — sub-detectable damage is assumed present but proven non-propagating under spectrum loading. FAA AC 20-107B governs.

14. Repair effectiveness. A perfectly executed scarf repair recovers ~85–90 % of pristine strength. A bolted patch typically 60–70 %. Composite repair quality is heavily technician-dependent and lacks the standardised metallurgical predictability of welding.

---

11. Cross-references

• [[Engineering/materials-steel]] — sibling structural material; metals vs composites cost / weight trade
• [[Engineering/materials-aluminum]] — sibling structural material; main aerospace alternative
• [[Engineering/materials-polymers]] — matrix chemistry and unreinforced polymer baselines
• [[Engineering/materials-selection]] — Ashby method comparing composites with metals and polymers
• [[Engineering/mechanics-of-materials]] — stress-strain foundations CLT builds on
• [[Engineering/structural-analysis]] — composite-stiffened panel buckling, plate bending
• [[Engineering/joining-welding]] — adhesive bonding for composites; thermoplastic welding
• [[Engineering/additive-manufacturing]] — continuous-fibre 3D printing of composites
• [[Engineering/aerodynamics]] — modern composite airframe structural drivers
• [[Engineering/fatigue-analysis]] — composite fatigue and damage tolerance
• [[Robotics/manipulator-design]] — CFRP robot links, end-effector tubes
• [[Robotics/multirotor-design]] — drone CFRP airframes and motor arms
• [[Languages/Tier3/construction-bim]] — STEP material assignments for laminates (AP242 composites extension)

---

12. Citations

1. Jones, R. M. Mechanics of Composite Materials, 2nd ed. (Taylor & Francis, 1999). The standard CLT textbook.
2. Daniel, I. M. & Ishai, O. Engineering Mechanics of Composite Materials, 2nd ed. (Oxford University Press, 2006).
3. Tsai, S. W. Theory of Composites Design (Think Composites, 1992). The Tsai-Wu criterion and laminate optimisation.
4. Hull, D. & Clyne, T. W. An Introduction to Composite Materials, 2nd ed. (Cambridge University Press, 1996).
5. Mallick, P. K. Fiber-Reinforced Composites: Materials, Manufacturing, and Design, 4th ed. (CRC Press, 2023).
6. Mazumdar, S. K. Composites Manufacturing: Materials, Product, and Process Engineering (CRC Press, 2002).
7. Composite Materials Handbook CMH-17 (former MIL-HDBK-17). Six-volume set: Vol 1 (polymer matrix, materials usage), Vol 2 (polymer matrix, materials properties — B-basis allowables), Vol 3 (polymer matrix, materials usage, design), Vol 5 (ceramic matrix), Vol 6 (structural sandwich). SAE International, current revision.
8. ASTM D3039 / D3039M-17 — Standard Test Method for Tensile Properties of Polymer Matrix Composite Materials.
9. ASTM D3410 / D3410M-16 — Standard Test Method for Compressive Properties of Polymer Matrix Composite Materials with Unsupported Gage Section.
10. ASTM D3518 / D3518M-18 — Standard Test Method for In-Plane Shear Response by Tensile Test of a ±45° Laminate.
11. ASTM D6671 / D6671M-22 — Mixed Mode I–Mode II Interlaminar Fracture Toughness of Unidirectional Fiber Reinforced Polymer Matrix Composites.
12. ASTM D7137 / D7137M-17 — Compressive Residual Strength Properties of Damaged Polymer Matrix Composite Plates (CAI).
13. ASTM D5528 / D5528M-21 — Mode I Interlaminar Fracture Toughness of Unidirectional Fiber-Reinforced Polymer Matrix Composites.
14. ASTM D5766 / D5766M-11 — Open Hole Tensile Strength of Polymer Matrix Composite Laminates.
15. ASTM D5961 / D5961M-17 — Bearing Response of Polymer Matrix Composite Laminates.
16. ISO 14130:1997 — Determination of apparent interlaminar shear strength by short-beam method.
17. FAA AC 20-107B — Composite Aircraft Structure (compliance and certification guidance).
18. FAA AC 20-53B — Protection of Aircraft Fuel Systems against Fuel Vapor Ignition due to Lightning.
19. Toray Composite Materials America. T700S, T800S, T1100G technical data sheets (current rev).
20. Hexcel Corporation. HexPly 8552 epoxy matrix product data sheet.
21. Solvay Composite Materials. Cycom 977-3 and Cycom 5250-4 BMI product datasheets.